Overview
Save Big on Coursera Plus. 7,000+ courses at $160 off. Limited Time Only!
Explore a 58-minute mathematics lecture that delves into Log-Sobolev inequalities (LSIs) for matrix-valued functions and their applications across probability theory, statistical mechanics, and information theory. Learn about the significant dichotomy between standard $L_2$-LSI and modified log-Sobolev inequality, where the former fails in matrix-valued settings while the $L_1$ variant maintains validity for matrix-valued functions in numerous important examples. Discover the implications of these findings, particularly in relation to Gaussian-type concentration inequalities for random matrices and quantum Markov semigroups.
Syllabus
Li Gao: Log-Sobolev inequalities for matrix-valued functions
Taught by
Hausdorff Center for Mathematics